University of California, San Francisco. Susan. Stewart is with the Northern California Cancer. Center, Union City. Requests for reprints should be sent to.
The Effectiveness of a Media-Led Intervention to Reduce Smoking among Vietnamese-American Men Christopher N. H. Jenkins, MA, MPH, Stephen J. McPhee, MD, Anh Le, Giao Qui Pham, MD, MS, Ngoc-The Ha, MSW, and Susan Stewart, PhD
Introduction
.......
Since the first health warnings about smoking were issued more than 30 years ago, smoking prevalence in the United States has declined from a high of 42% in 1965 to 26% in 1994.1 Among Vietnamese-American men, however, smoking prevalence rates remain stubbomly high; rates ranging from 35% to 56% have been reported in the literature.2- Since the end of the war in Vietnam in 1975, the population of Vietnamese in the United States has grown to more than 600 000.7 Nearly half (46%) have settled in California. This paper reports the effectiveness of an intervention designed to lower smoking prevalence among Vietnamese men in the San Francisco Bay area. The intervention was set in the context of a statewide tobacco control initiative that included an increased cigarette excise tax, a mass education campaign, and new ordinances restricting smoking.'°0 From 1990 to 1992 in San Francisco and Alameda counties, we mounted a 2-year media-led campaign that we evaluated in a controlled trial. This intervention was preceded by an uncontrolled, 15month, pilot anti-tobacco media program. Anti-tobacco activities were conducted, therefore, for a total of 39 months. Here we present the results of an evaluation of the effectiveness of the 2-year intervention. Since our earlier research showed that recent male Vietnamese immigrants were more likely to smoke, the campaign targeted this subpopulation. Houston, Tex, served as the comparison community. Women were excluded from the study because of their negligible smoking prevalence rates.2-5
Methods We published articles in Vietnameselanguage newspapers resulting in 465 000 print media exposures. We produced a Vietnamese-language videotape, which was broadcast twice on Vietnameselanguage television, and several Vietnamese-language health education materials,
including a calendar, bumper stickers, lapel buttons, three posters, two brochures, and a quit kit. More than 15 000 copies of the brochures and 4600 copies of the quit kit were distributed. We also conducted an anti-tobacco Vietnamese-language counteradvertising campaign that included three different billboards, newspaper advertisements resulting in nearly 2.8 million exposures, and paid television advertisements that accounted for more than 100 minutes of air time. Project staff delivered presentations at 25 community events. Because a majority of Vietnamese visit Vietnamese physicians,2 we organized a continuing medical education course on smoking cessation counseling methods for 68 Vietnamese physicians. In two Vietnamese-language "Saturday" schools, we organized anti-tobacco activities for a total of 400 students.'2 Finally, we distributed Vietnamese-language "no smoking" signs and smoking control ordinances to Vietnamese restaurants and businesses. We hypothesized that the reduction in smoking prevalence following the intervention and the proportion of smokers who quit during the intervention would be significantly greater in the experimental than in the comparison community. Using a quasi-experimental control group design,'3 we conducted cross-sectional telephone surveys of independent samples of approximately 1200 randomly selected Vietnamese men in Christopher N. H. Jenkins, Stephen J. McPhee, Anh Le, Giao Qui Pham, and Ngoc-The Ha are with Suc Khoe La Vang! (Health is Gold!), Vietnamese Community Health Promotion Project, Division of General Internal Medicine, Department of Medicine, University of California, San Francisco. Stephen J. McPhee is also with the Institute for Health Policy Studies, University of California, San Francisco. Susan Stewart is with the Northern California Cancer Center, Union City. Requests for reprints should be sent to Christopher N. H. Jenkins, MA, MPH, Vietnamese Community Health Promotion Project, Division of General Internal Medicine, University of California, San Francisco, 44 Montgomery, Suite 850, San Francisco, CA 94104. This paper was accepted August 21, 1996.
.......
American Journal of Public Health 1031 ...
......
Public Health Briefs
smokers if they responded no to the first question, as former smokers if they responded yes to the first question and no to the second question, and as current smokers if they responded yes to both questions. Current smokers were asked amount and duration of smoking, how much they wanted to quit smoking, and how difficult they thought it would be to quit; they were also asked whether they had ever tried to quit and, if so, on how many occasions. Former smokers were asked when they had quit smoking. We asked former and current smokers whether they had ever been advised by a physician to quit. We assessed exposure to intervention activities by asking respondents whether they had ever read, seen, or attended any of the five elements of the media intervention. Demographic data collected included age, year of immigration to the United States, education, English-language proficiency, employment status, and poverty status (according to federal criteria'5"16). To test our hypotheses, we examined two outcomes: (1) the proportion of current smokers and (2) the proportion who had quit smoking during the 2 years prior to either the pretest or posttest interview. Analyses described differences in outcomes between experimental and comparison subjects at pretest and posttest. In preliminary analyses of sociodemographic characteristics and smoking behaviors, we used t tests to examine the significance of differences in means and chi-square tests to examine the significance of differences in proportions. Using the LOGISTIC procedure in the SAS statistical package,'7 we performed multiple logistic regression analyses to control simultaneously for differences in sociodemographic factors that might have accounted for differences in outcomes. We computed models on the pooled data to identify variables significantly associated with non-smoking and quitting during the prior 2 years, the major dependent variables. The analyses involved three dummy variables: time (O pretest, 1 posttest), to assess secular trends in the control community; site (O comparison, 1 = intervention), to assess site differences at baseline; and Time X Site, the intervention term (1 = intervention site at posttest, 0 = all other conditions), to assess the intervention effect (i.e., the difference between the two communities in the change over time). Terms that were not significant (at P < .05) were successively eliminated to obtain parsimonious models. Adjusted odds ranever
TABLE 1 -Demographic Characteristics of Respondents to Pretest and Posttest Telephone Surveys: Comparison (Houston) and Intervention (San Francisco) Communities
Posttest (1992)
Pretest (1990)
San San Francisco Houston Francisco Houston (n = 1581) (n = 1133) (n = 1209) (n = 1202) 38.2 71 35 13 6 1979
Mean age, y Limited English proficiency,c % High school education or less, % Below poverty level, % Unemployed, % Mean year of immigration
38.7
80d 55d 39d 18d 1981d
41.7a 77a 41 a 26a 1oa 1981a
38.8b 84b,e 57b 43b,e 27be
1983be
aSignificant 1990/1992 difference (P < .0003). bSignificant Houston/San Francisco difference (P < .0002). CNone, poor, or fair (vs good or fluent). dSignificant Houston/San Francisco difference (P < .0001). eSignificant 1990/1992 difference (P < .05).
TABLE 2-Smoking and Quitting Behavior of Vietnamese Male Respondents to Pretest and Posttest Telephone Surveys: Comparison (Houston) and Intervention (San Francisco) Communities Pretest (1990)
Posttest (1992)
San San Francisco Houston Francisco Houston (n = 1581) (n = 1133) (n = 1209) (n = 1202) All respondents Current smokers, % Quit during prior 2 years, % Ever smoked, % Current smokers No. Mean no. cigarettes per
39.6 5.8 54.0 626 13.2 (8.7)
day (SD) Definitely wants to quit, % Thinks quitting is very difficult, % Ever tried to quit, % Mean no. quit attempts (SD) Current and former smokers No. Ever advised by doctor to quit, %
36.1 7.2 53.6
40.9 7.4 63.7c
33.9a
409
494
407
11.1d (8.2) 11.9c (8.0) 10.3a (7.1)
23 25
29 29
44c 26
49 1.1 (1.4)
61 d 1.4 (1.4)
77c 2.5c (6.6)
817 12
1 0.2b
57.5a
592 23
706 39c
45b 33
73b 1.8 (2.1) 638
50b
aComparison of posttest Houston and posttest San Francisco: P . .01. bComparison of pretest San Francisco and posttest San Francisco: P s .01. COomparison of pretest Houston and posttest Houston: P .01. dComparison of pretest Houston and pretest San Francisco: P s .01. -
=
each community at both preintervention and postintervention. Survey subjects were required to be male, 18 years of age or older, and Vietnamese speaking. The sampling frame consisted of telephone numbers chosen randomly from the 23 most common Vietnamese sumames listed in area telephone books. After enumerating all eligible men, survey workers selected a subject for interview according to random 1032 American Journal of Public Health
age rankings in the household.'4 The University of California, San Francisco, Committee on Human Research approved the research protocol. The survey instrument was developed in English, translated into Vietnamese, back-translated, and pilot tested. Subjects were asked whether they had ever smoked cigarettes and whether, during the prior week, they had smoked a cigarette. Respondents were classified as
=
=
June 1997, Vol. 87, No. 6
Public Health Briefs
tios (ORs) were computed for covariates with 95% confidence intervals (CIs).
Results Response rates were similar for the intervention and comparison communities (pretest: 84% in San Francisco and 82% in Houston; posttest: 94% in San Francisco and 88% in Houston). (In calculating response rates, we excluded call attempts that reached businesses, answering machines, disconnected telephones, households without eligible men, and telephones with no answer despite five call-back attempts at different times of day.) There were significant differences among the four samples on nearly all sociodemographic characteristics (Table 1). These variables were included, therefore, in the multivariate analyses. As expected, recall of each of the five elements of the media intervention was significantly greater in San Francisco than in Houston (P < .05), with the exception of newspaper articles. In both San Francisco (P < .01) and Houston (P < .01), smokers were more likely to recall the elements than non-smokers. At pretest, the smoking prevalence rate was 36.1% in San Francisco, somewhat lower than the 39.6% rate in Houston (Table 2). Between pretest and posttest, smoking declined in San Francisco, while the rate in Houston increased, resulting in a net change of -3.5 percentage points. The posttest smoking rate in San Francisco was significantly lower than the rate in Houston (P = .004). The rate of quitting during the prior 2 years, which rose in both communities, rose more steeply in San Francisco, resulting in a net change of 1.4 percentage points. At post-test, the quitting rate was higher in San Francisco than in Houston (P = .0 17). Current smokers in San Francisco smoked significantly fewer cigarettes per day than those in Houston, both at pretest and at posttest. At posttest, motivation to quit smoking and the rate of ever trying to quit improved significantly among smokers in both communities. There was no significant change in either community in terms of how difficult smokers felt quitting would be. Current smokers in Houston reported a significantly greater number of quit attempts at posttest, while there was no change in this behavior among San Francisco smokers. Finally, at posttest, ever smokers in both communities with regular physicians were more likely than those without regular physicians to have June 1997, Vol. 87, No. 6
TABLE 3-Multiple Logistic Regression Analyses of Predictors of Current Smoking among Vietnamese Men: Comparison (Houston) and Intervention (San Francisco) Communities (n = 4873) Adjusted Odds Ratio
95% Confidence Interval
1.00 0.82
Referent 0.68, 0.99
1.00 0.75
Referent 0.64, 0.87
0.54 1.00 0.51 0.36 0.32
0.42, 0.68 Referent 0.43, 0.60 0.28, 0.46 0.23, 0.44
1.00 0.69
Referent 0.58, 0.82
1.00 0.75 0.50 1.05
Referent 0.63, 0.90 0.36, 0.70 1.03, 1.06
1.00 0.47 1.24
Referent 0.37, 0.61 1.03, 1.48
Intervention term Houston posttest/pretest San Francisco posttest/pretest Site Pretest: Houston Pretest: San Francisco Age, y 18-24
25-44 45-54 55-64 -65 Education Less than college College or more English-language proficiency None/fair/poor Good Fluent Immigration: each additional calendar year (more recent immigrants) Employment
Employed/disabled/retired Student Unemployed
Note. In a full model, time, poverty status, and number of antismoking media elements recalled were not significant terms at the .05 level and were therefore dropped from further analyses to develop this parsimonious model.
been advised by a physician to quit smoking. After control for baseline differences in smoking prevalence between the two sites and demographic differences among the four samples, the odds of being a smoker at postintervention in San Francisco were significantly lower than in Houston, indicating that the intervention was effective in reducing smoking prevalence (Table 3). The odds of being a smoker at postintervention were lower for respondents who were 18 to 24 years of age or 45 years of age or older than for those who were 25 to 44 years old; odds were also lower for respondents who had at least a college education, those with good or fluent English-language proficiency, and students. The odds of being a smoker were greater for more recent immigrants and the unemployed. At postintervention, the increase in the odds of quitting during the prior 2 years was significantly higher in San Francisco than in Houston (OR = 1.65, 95% CI = 1.27, 2.15), indicating that the
intervention was effective in increasing quitting during the intervention in San Francisco. The strongest predictor of quitting was being a student (OR 2.19, 95% CI 1.45, 3.33); other predictors included more recent year of immigration (OR 1.03, 95% CI 1.00, 1.05), each additional year of age (OR 1.03, 95% CI 1.02, 1.04), and having at least a high school education (OR 1.33, 95% CI 1.04, 1.70). =
=
=
=
=
=
=
=
Discussion The results reported here indicate modest success of an antismoking intervention directed at Vietnamese men. The intervention produced both a significant decrease in smoking prevalence and a significant increase in quitting among men in the intervention community relative to men in the comparison community. In comparison with earlier studies, the smoking prevalence rates for Vietnamese men in San Francisco shown in Table 2 indicate an encouraging downward trend; American Journal of Public Health 1033
Public Health Briefs
data gathered in 1987 showed a rate of 56%.2 The predictors of cigarette smoking identified here (less education, less English-language proficiency, more recent immigration) remain generally consistent with results of earlier studies.26 These results, however, contrast with findings from a similar, contemporaneous campaign, reported previously, that showed no effect.6 Why did the intervention reported here show an effect, while a similar intervention targeting neighboring Santa Clara County showed none? The difference may be explained in two ways. First, the San Francisco intervention was carried out for a longer period of time (39 months vs 24 months in Santa Clara). Others have shown that the duration of a mass media-led antismoking campaign may be a critical factor in achieving success. 8-24 Second, the San Francisco intervention had an additional component targeting students and their families. It is possible that some of the changes observed between 1990 and 1992 occurred as a result of "spillover" from the statewide campaign and the general downward trend in smoking among male Californians during this period (from 25.5% to 22.8%25). However, the limited English-language proficiency of the Vietnamese suggests that the state's primarily English-language campaign may have had little impact on this population. Our results, on the other hand, show that the Vietnamese-language intervention achieved broad penetration in the target community, especially among smokers. Several limitations are inherent to this study: use of cross-sectional samples rather than a cohort design, restriction of sampling frames to households with listed telephone numbers, and possible lack of generalizability to Vietnamese elsewhere. In addition, the study design was limited to two communities, precluding randomization of communities (or individuals) to experimental or control conditions. Measurements were limited to two points in time, raising the possibility that results were due simply to chance. There is internal consistency in the targeted behaviors, however, and the observed changes make it likely that the intervention caused the decline in smoking rates in San Francisco. Furthermore, the design might have been strengthened by adding another comparison community within California in order to detect any additional differences due to exposure to our intervention rather than to the larger statewide campaign alone. Finally, we did not biochemically validate smoking status. One report 1034 American Journal of Public Health
has shown, however, that Vietnamese male smoking prevalence estimates derived from self-reports do not differ from estimates obtained by saliva cotinine assay.26 Despite the success reported here, smoking prevalence among Vietnamese men remains alarmingly high. The postintervention rate of 34% in San Francisco remains more than 1.5 times the year 2000 goal of 20% set by the US Public Health Service for Southeast Asian men.27 Further progress in tobacco control among Vietnamese will require sustained, multidimensional efforts, including both broad policy initiatives targeting the general population and culturally tailored, language-appropriate campaigns aimed specifically at the Vietnamese population. D
11.
12.
13.
14.
15. 16.
Acknowledgments This research was supported by funds provided by the Cigarette and Tobacco Surtax Fund of the State of California through the Tobacco Control Section (contract 90-10984). These results were presented in part at the 9th World Conference on Tobacco and Health, October 1994, Paris, France, and at the annual meeting of the Society of General Internal Medicine, May 1995, San Diego, Calif.
17.
18. 19.
References 1. Centers for Disease Control and Prevention. Cigarette smoking among adultsUnited States, 1994. MMWR Morb Mortal Wklv Rep. 1996;45:588-590. 2. Jenkins CNH, McPhee SJ, Bird JA, Bonilla N-T. Cancer risks and prevention practices among Vietnamese refugees. West J Med. 1990; 153:34-39. 3. Centers for Disease Control and Prevention. Behavioral risk factor survey of Vietnamese in California-1991. MMWR Morb Mortal Wklv Rep. 1992;41 :69-72. 4. Centers for Disease Control and Prevention. Cigarette smoking among Chinese, Vietnamese, and Hispanics-California, 1989-199 1. MMWR Morb Mortal Wklv Rep. 1992;41 :362-367. 5. Jenkins CNH, McPhee SJ, Bonilla N-T, Nam TV, Chen A. Cigarette smoking among Vietnamese immigrants in California. Am J Healtk Promotion. 1995;9:254-256. 6. McPhee SJ, Jenkins CNH, Wong C, et al. Smoking cessation intervention among Vietnamese Americans: a controlled trial. Tobacco Control. 1995;4(suppl 1):S 16-S24. 7. We the American ... Asians. Washington, DC: US Dept of Commerce. Bureau of the Census; 1993. 8. Pierce JP, Burns DM, Berry C, et al. Reducing tobacco consumption in California: Proposition 99 seems to work. JAMA. 1991 ;265: 1257-1258. 9. Popham WJ, Potter LD, Bal DG, Johnson MD, Duerr JM, Quinn V. Do anti-smoking media campaigns help smokers quit'? Public Health Rep. 1993; 108:510-513. 10. Hu T-W, KeelerTE, Sung H-Y, Barnett PG. The impact of California anti-smoking legislation on cigarette sales, consumption
20.
21.
22.
23.
24.
25.
26.
27.
and prices. Tobacco Control. 1995;4(suppl 1 ):S34-S38. McPhee SJ, Bird JA, Ha N-T, Jenkins CNH, Fordham D, Le B. Pathways to early cancer detection for Vietnamese: Suc Khoe La Vang! (Health is Gold!). Health Educ Q. 1996;23(suppl):S60-S75. Le A, McPhee SJ, Jenkins CNH, Fordham D, Ha N-T, Pham GQ. Tobacco prevention and cessation among Vietnamese: youth as couriers of new information. Abstract presented at the 122nd Annual Meeting of the American Public Health Association; October 31, 1994; Washington, DC. Cook TD, Campbell DT. Quasi-E&perimenitation: Design and Analysis Issues for Field Studies. Chicago, III: Rand McNally College Publishing Co; 1979:115-117. Kish L. Survey Sampling. New York, NY: John Wiley & Sons nc; 1965. Annual update of the federal poverty income guidelines. Federal Register February 16, 1990;55:5664-5666. Income eligibility guidelines, July 1, 1992June 30, 1993. Federal Register March 25, 1992; 57(58):1033 1. SAS Technical Report P-200, SAS/STAT Software: CALIS and LOGISTIC Procedures, Release 6.04. Cary, NC: SAS Institute Inc; 1990. Flay BR. Mass media and smoking cessation: a critical review. Am J Public Health. 1987;77: 153-160. Farquhar JW, Fortmann SP, Flora JA, et al. Effects of communitywide education on cardiovascular disease risk factors, the Stanford Five-City Project. JAMA. 1990; 264:359-365. Salonen JT, Puska P, Kottke TE, Tuomilehto J. Changes in smoking, serum cholesterol, and blood pressure levels during a community-based cardiovascular disease prevention program-the North Karelia Project. Am J Epidemiol. 1981; 114:81-94. Fortmann SP, Taylor CB, Flora JA, Jatulis DE. Changes in adult cigarette smoking prevalence after 5 years of community health education: the Stanford Five-City Project. Am J Epidemiol. 1993; 137:82-96. COMMIT Research Group. Community Intervention Trial for Smoking Cessation (COMMIT): I. Cohort results from a four-year community intervention. Am J Public Health. 1995;85:183-192. COMMIT Research Group. Community Intervention Trial for Smoking Cessation (COMMIT): II. Changes in adult smoking prevalence. Am J Public Health. 1995;85: 193-200. Susser M. Editorial: The tribulations of trials-intervention in communities. Am J Public Health. 1995;85:156-158. Pierce JP, Farkas A, Evans N, et al. Tobacco Use in California 1992: A Focus on Preventing Uptake in Adolescents. Sacramento, Calif: California Dept of Health Services; 1993. Wewers ME, Dhatt RK, Moeschberger ML, Guthrie RM, Kuun P, Chen MS. Misclassification of smoking status among Southeast Asian adult immigrants. Am J Respir Crit Care Med. 1995;152:1917-1921. Public Health Service. Healthy People 2000: National Health Promotionl and Disease Prevention Objectiv!es. Washington, DC: US Dept of Health and Human Services; 1991.
June 1997, Vol. 87, No. 6
